如何在Python 3.7的DataFrame中按条件替换列值？

Got it, let's break down how to implement your two replacement rules efficiently in pandas. Here's a step-by-step approach that matches exactly what you need:

Step 1: Set up your sample DataFrame #

First, let's recreate your input data so you can test the code directly:

import pandas as pd

# Sample DataFrame
data = {
    'XX': [0, 1, 3, -1, 5, 7, -1, 6],
    'Date': ['2016-05-01']*8,
    'Time': ['19:00:00', '18:00:00', '17:00:00', '16:00:00', 
             '15:00:00', '14:00:00', '13:00:00', '12:00:00']
}
df = pd.DataFrame(data)

# Your lookup dictionary (included the 13:00 entry as seen in your example)
lookup_dict = {'13:00:00': 1, '01:00:00':1, '02:00:00':4, '23:00:00':0}

Step 2: Apply the specific conditional replacement first #

We want to handle the special case first (XX=-1 and Time='16:00:00') to make sure it doesn't get overwritten by the dictionary mapping later. Use pandas loc to target exactly those rows:

# Replace XX with 2 where conditions are met
df.loc[(df['XX'] == -1) & (df['Time'] == '16:00:00'), 'XX'] = 2

Step 3: Use the lookup dictionary for remaining -1 values #

Now handle all other rows where XX is still -1, mapping their Time values to the corresponding value in your dictionary:

# Replace remaining XX=-1 entries using the lookup dict
df.loc[df['XX'] == -1, 'XX'] = df.loc[df['XX'] == -1, 'Time'].map(lookup_dict)

Step 4: Verify the result #

If you print the DataFrame now with print(df), you'll get exactly the output you wanted:

XX        Date      Time
0   0  2016-05-01  19:00:00
1   1  2016-05-01  18:00:00
2   3  2016-05-01  17:00:00
3   2  2016-05-01  16:00:00
4   5  2016-05-01  15:00:00
5   7  2016-05-01  14:00:00
6   1  2016-05-01  13:00:00
7   6  2016-05-01  12:00:00

Key Notes: #

• Order matters: We handle the special case first because if we used the dictionary mapping first, the 16:00:00 entry would be replaced with whatever value is in your dict (if any) instead of 2.
• loc is safe: Using loc ensures we only modify the exact rows/columns we intend to, avoiding accidental changes to other data.
• map is efficient: It quickly matches Time values to your dictionary, which is perfect for bulk replacements.

内容的提问来源于stack exchange，提问作者Clueless_Doggo